The OpenOffice fork is officially here (Computerworld)

I am not officially involved in the fork. Let me try to address your concerns: 1) OO is now fragmented and 2) it's like Iceweasel and Firefox

1) What do you want to do when upstream who controls commit access is decreasing the resources it puts into the project and at the same time does not take in contributions by others. Or rejects them just in order to re-code the same stuff with its own resources. And who requires full copyright assignments so they can sell your code in their proprietary version. What can you do about it? Go-oo kept a bunch of patches on top of the code, but that becomes unmanageable, as the number of applied patches was already around 1,000 or so. Other distros also kept their own patches, like redhat, suse, etc. I would argue that OOo WAS already fragmented as each distro did their micro-improvements in their isolated islands. What Libreoffice does is to collect all these patches, reducing fragmentation. They did invite Oracle to join the fun, and Oracle said no (for now). So what can you really do when you want to see improvements, rather than stagnation in the code? Currently there are code cleanups going on of code that has been written in 1997.

2) Iceweasel vs Firefox. It was hoped that Oracle donated the name when joining, so we could all be happy with the OOo (what a name!!!) that we know. They did not, and I say it is a different thing that rebranding due to trademark policy. This is a full fledged fork now. They try to merge in OOo improvments, but as the code diverges this is going to be harder and harder. So this is not the same old wine in a new Iceweasel-bottle, this is (going to become) a different beast.

Is it optimal? Heck no, forks are always painful and a loss of productivity in the short run. But there are plenty of examples where forks turned out to spurn innovation and development in the long run.